OBJECTIVE: To assess the strength of association between alternative measures of electroencephalographic (EEG) signal peak-to-peak amplitude (ppA) and postmenstrual age (PMA) among a cohort of extremely premature infants.
METHODS: 177 Two-channel EEG recordings 3-6h long were collected from 26 infants born before 29weeks of gestation. The raw EEG was converted into four different continuous measures of ppA: amplitude-integrated EEG (aEEG), range-EEG (rEEG), Gotman and Gloor's half-wave decomposition (HWD), and root of mean squares (RMS). For each ppA-measure EEG indices including mean, median, and 5% margins; indices of spread (width, standard deviation, coefficient of variation), and asymmetry were calculated for each 1min epoch. The medians of each index for the entire recording were tested for association with PMA using linear mixed models.
RESULTS: The log-transformed values of aEEG and rEEG indices of spread were highly associated with PMA (fixed effects R(β)(2)=0.84-0.89).
CONCLUSIONS: Indices of spread by aEEG or rEEG can be used as indicators of neonatal brain maturation. However, rEEG produces the absolute values that most closely approximate the raw EEG amplitudes.
SIGNIFICANCE: The indices of spread and rEEG as a measure of ppA provide a basis for improvements in neonatal EEG monitoring.